INTEGRATING TEXT-CNN AND TOPIC MODELINGFOR DARK WEB ANALYSIS

Abstract

The Dark Web's safety is contingent upon content analysis, which enables anonymous communication and promotes illegal activities. Research on Topics and TextCNN Weights are used in this inquiry to apply classification systems. Patterns may be extracted from Dark Web pages using topic modeling techniques like LDA and NMF. The extracted themes improve Text-CNN's ability to categorize cybercrime, extremism, and fraud. The suggested approach improves Dark Web surveillance in terms of accuracy and scalability. Experiments show that adding topic modeling improves classification effectiveness. New dangers can be detected by law enforcement using this technique. Innovative ideas are supplied to those working in the protection industry. By mining the Dark Web for relevant content, the approach is advanced.